Re: CCL (Was IESG Information architecture)

LYNCH@jade.bucknell.edu Mon, 28 September 1992 22:10 UTC

Date: Mon, 28 Sep 1992 16:47:24 -0400
Sender: ietf-archive-request@IETF.NRI.Reston.VA.US
From: LYNCH@jade.bucknell.edu
Message-Id: <920928164724.2069@jade.bucknell.edu>
Subject: Re: CCL (Was IESG Information architecture)
To: NIR@bunyip.com

>I'm starting to remember the context; as I remember, the problem with CCL
>in the context of Z39.50 and automatic usage is that it's big on verbs,
>but short on nouns- it doesn't have enough semantic specification. If this
>is wrong, I'd like to know (It's not something I've ever seen.
>
>Simon

As I said I'm no expert.  Our CCL-compliant system includes author, title,
subject, publisher, notes, and several other qualifiers.  I'm not
sure if these are specified by the standard or not.  I believe the 
standard says you must use AU if you are going to allow qualification 
by author, TI to qualify by title etc.  But it doesn't mandate
that you allow qualification by anything, nor does it specify, for
example, which MARC tags must be included in an AU search.  Simply,
if you are going to provide an author index (inverted file) and you
want to allow searches of it you need to use AU.  Of course there
may be things by which you want to qualify for which no mnemonic has
been specified.  But you can always allow unqualified searches that
are essentially keyword searches of whatever parts of the record
have been keyword indexed.

Mike

Re: CCL (Was IESG Information architecture) LYNCH